Hi all you happy, shiney people.
I just learned my friend (who I almost never see), Ernie, is going to Japan for six months as he was accepted into a highly desirable language program. Congrats to Ernie! The downside is that slowass.net etc are hosted from his closet on his T1, and the NetBSD macppc machine (the old 7300) crashes slightly more frequently than that. I seriously don't think Ernie will give me a key.
So, I have to start migrating stuff to my home network. Network Solution's create-nameserver utility is as impossible to find as ever. https://www.networksolutions.com/en_US/manage-it/manage-nameservers.jhtml ... that's the magic address, for after you're logged in.
At home, I've got on the network my Toughbook CF-R1 with its broken screen and a monitor as my primary workstation, my Toughbook CF-28, the Compaq Aeo 4/25 notebook, a Via hooked up the TV, and for servers, a Mac 7300 that's supposed to be a spare for the Mac serving slowass.net, and now a freshly built Sparc 10 (non-Ultra). I've ordered a replacement mainboard for my DEC3100. I couldn't get serial console on it before and the solid 8 LEDs on the back indicate some kind of serious problem. Let's see if this fixes it. That'll be one more server. I just got a new hub with my last paycheck, a bookshelf to help organize this mess, some patch cords, etc.
Got djbdns running on the home Mac, which is the home fileserver. The SOA record type automatically creates A records for the primary DNS, as does the NS record, which causes duplicate A records to be created -- and served -- when several domains are hosted from one tinydns/djbdns server. For 8 domains, it's annoying, but an ISP with thousands would find this unusable -- unless there's a work-around I'm missing. Aside from being kind of cranky software, it doesn't give me the seriously bad vibe that bind does. Bind seems to have completely and utterly not thought anything out -- things that should be serious aren't and vice versa, under moderate load it pegs the CPU (absurd considering how very little DNS has to actually do), it has a long history of flaws (again, absurd considering how little it actually has to do). djbdns/tinydns at least actually inspires some confidence.
I put perldesignpatterns.com on registration -- manual registration. Email me if you want the password. This uber-sucks, I know. The spammers won. Everything I've thrown at them, they've adapted to. It's all automated and re-automated and distributed across countless comprimised machines and open proxies. PDP has been dying a slow death of neglect anyway. I'd love to go back in and rework it all and fix it up, but the Perl community itself is atrophying, with knowledgable programmers jumping ship rapidly and novices tending towards PHP. So, I'm going to have to start seriously forcing myself to re-evaluate community efforts.
Perl 6 Now wasn't a bestseller... in about a year, Amazon had dropped it, then picked it back up again. It didn't recoup. But for some reason, I've gone and registered perlsecurity.com, and... well... let's see. You'd think I'd know better.
That's my report on the sysadministration front.
-scott
Re:Really? Atrophying?
TeeJay on 2006-06-04T18:19:46
I don't see it myself.
I see a small number of talented developers broadening their knowledge but none abandoning perl. You'd be a complete fool to discard a tool that works well and with which you've become an expert, even if you found some other tools that work nicely.Re:Really? Atrophying?
scrottie on 2006-06-14T02:55:41
Oh yes, silly me. Market forces don't exist. I must be on crack. Programmers use whatever tool they think best. It would be a horrible world if we had to do things like walk into jobs where the language was already speced out by CTOs and project managers who based their decisions partially on following fads -- both towards technologies that get a lot of media and away from technologies that get a bad rap for being messy, unmaintainable, and having generally having been used by novices.
Go talk to a recruiter. Ask him or her how many Perl jobs she has and then ask her how many Java jobs she has. Then ask how many C# jobs she has -- not that that many more people use C# but there are a hell of a lot of openings.Re:Really? Atrophying?
scrottie on 2006-06-14T02:57:05
Oh yes, silly me. Market forces don't exist. I must be on crack. Programmers use whatever tool they think best. It would be a horrible world if we had to do things like walk into jobs where the language was already speced out by CTOs and project managers who based their decisions partially on following fads -- both towards technologies that get a lot of media and away from technologies that get a bad rap for being messy, unmaintainable, and having generally having been used by novices.
Go talk to a recruiter. Ask him or her how many Perl jobs she has and then ask her how many Java jobs she has. Then ask how many C# jobs she has -- not that that many more people use C# but there are a hell of a lot of openings.
Re:Really? Atrophying?
TeeJay on 2006-06-15T09:37:15
Sorry this is bullshit.. there has always been more Java than perl jobs, this is because Sun and Java product vendors spent a hell of a lot of money to make that happen, and still do.
In fact I thought I'd post on my journal a proper response with actual facts : http://use.perl.org/user/TeeJay/journal/29915Re:Really? Atrophying?
scrottie on 2006-06-14T02:51:45
ORLY?
